We have demonstrated that immunostaining with a mixture of monoclonal antibodies that react with neuroblastoma but not normal marrow cells can detect one tumor cell per 100,000 normal cells and that immunocytology is more sensitive than conventional smears and trephine biopsies for detecting metastases in marrow. In 92 cases, we have demonstrated that immunocytology provides a highly sensitive means of evaluating the efficacy of ex vivo purging of tumor cells from autologous marrow. Our recent study demonstrated that immunocytologic analysis of marrow at diagnosis provides clinically useful prognostic information. The Childrens Cancer Study Group (CCSG) has protocols for all newly diagnosed patients with neuroblastoma that stratify therapy according to clinical stage, age, N-myc gene copy number and expression, histopathology (Shimada), and serum ferritin; additional prognostic factors are being studied and added as indicated. Entry of approximately 150 newly diagnosed patients annually into these Group-wide studies wt glow us to continue integrated biologic, diagnostic, and therapeutic investigations. Immunocytologic analysis of bone marrow is an integral and essential part of these nationwide studies. The aims of the proposed research, which will be performed in the CCSG Neuroblastoma Biology Research Laboratory at Childrens Hospital of Los Angeles, are as follows. 1) Assess the efficacy of ex vivo purging of tumor cells from marrow used for autologous bone marrow transplantation (ABMT). Intensive chemoradiotherapy followed by ABMT is a promising therapy for poor prognosis patients that will be tested in a nationwide randomized trial by the CCSG. Autologous marrow will be "purged' of tumor cells ex vivo, and the efficacy of this will be assessed with immunocytology. The absence of detectable tumor cells will minimize the chance of infusing viable tumor cells with autologous marrow. 2) Determine the prognostic importance of marrow tumor cells before, during, and after therapy. We shall expand the number of cases studied at diagnosis so that the relative clinical importance of marrow tumor cells and other prognostic factors can be compared in multivariate analysis. We also shall quantitate the rate of response of marrow metastases to therapy and determine if this is a sensitive predictor of outcome and of efficacy for a given therapy regimen. 3) Determine the effect of biotherapeutic agents on marrow metastases. Immunocytology will be used to evaluate new biotherapeutic agents such as 13-cis-retinoic acid, IL-2, and monoclonal antibodies, which may have their greatest effect on microscopic residual disease (marrow metastases). We anticipate that these studies will contribute to a better understanding of neuroblastoma and to therapy that is more effective and specific.